Skip to content

Latest commit

 

History

History
199 lines (172 loc) · 7.79 KB

README.md

File metadata and controls

199 lines (172 loc) · 7.79 KB
Raw

Seattle Airbnb SQL Capstone Project

Introduction

This project link represents a detailed SQL analysis of Airbnb listings in Seattle, conducted as part of the Analytics Extra Data Analytics Mentorship Program. The analysis explores various aspects of the Airbnb market, including pricing trends, host revenue, guest sentiment, and seasonal trends. This project demonstrates my proficiency in SQL through data cleaning, transformation, aggregation, and insightful analysis.

Problem Statement

This analysis addressed the following business questions:

  1. Which neighborhood generates the highest total revenue, and what is the average listing price?
  2. How do average listing prices vary across neighborhoods?
  3. How frequently does the word "comfortable" appear in guest reviews?
  4. Which listings have a high proportion of positive reviews?
  5. How much revenue did each host generate from listings reviewed within the past year?
  6. How many hosts are named "Stan" (case-insensitive)?
  7. What is the total revenue from listings priced above $700?
  8. Which listing has the highest number of reviews?
  9. Which listings were last reviewed in the final quarter of 2022?
  10. How can listings be categorized as "high-priced" or "low-priced" based on neighborhood averages?

Skills and Concepts Demonstrated

  • SQL Joins: Inner, Left joins to combine data from multiple tables.
  • Data Cleaning: Type conversion, NULL value handling, and filtering.
  • Data Aggregation and Filtering: SUM, AVG, COUNT, and WHERE clauses.
  • Conditional Logic: Using CASE statements for categorization.
  • Window Functions and CTEs: Using ROW_NUMBER() and WITH clauses.

Data Overview

The dataset consists of three tables:

  1. Listings: Contains unique listing details, including location and pricing.
  2. Pricing: Captures pricing and availability data over time.
  3. Reviews: Stores guest reviews, ratings, and feedback for each listing.

Results and Analysis

1. Neighborhood with the Highest Revenue and Average Price

This query identifies the neighborhood with the highest total revenue and the average listing price.

Result: The neighborhood with the highest revenue and average price highlights areas of high demand.


SELECT TOP 1  l.neighbourhood,
    SUM(p.price) AS total_revenue,
    AVG(p.price) AS avg_price_per_listing
FROM listings l
JOIN pricing p ON l.id = p.listing_id
GROUP BY l.neighbourhood
ORDER BY total_revenue DESC;

2. Average Listing Price by Neighbourhood

Here, I compare the average price of listings across neighborhoods to identify pricing trends.

Result: Some neighborhoods have significantly higher average prices, indicating potential for premium markets.


SELECT 
    l.neighbourhood,
    AVG(p.price) AS avg_price
FROM listings 1
JOIN pricing p ON l.id = p.listing_id
GROUP BY l.neighbourhood
ORDER BY avg_price DESC;

3. Count of "Comfortable" in Reviews

This query counts the occurrences of the word comfortable in the reviews.

Result: The count shows guest sentiment trends based on commonly used terms.


SELECT 
    COUNT(*) AS comfortable_count
FROM reviews r
WHERE LOWER(CAST(r.comments AS VARCHAR (MAX))) LIKE '%comfortable%';

4. Listings with High Proportion of Positive Reviews

I filtered listings with positive review keywords (great, comfortable, good) to identify highly rated properties.

Result: Identifies listings that maintain a high ratio of positive feedback.


SELECT 
    l.id AS listing_id,
    l.name AS listing_name,
    COUNT(r.ID) AS total_reviews,
    SUM(CASE WHEN LOWER (CAST(r.comments AS VARCHAR (MAX))) LIKE '%great%' 
	OR LOWER(CAST(r.comments AS VARCHAR (MAX))) LIKE '%comfortable%' 
	OR LOWER(CAST(r.comments AS VARCHAR (MAX))) LIKE '%good%' THEN 1 ELSE 0 END) AS positive_reviews

FROM listings l
JOIN reviews r ON l.id = r.listing_id
GROUP BY l.id, l.name
HAVING (SUM(CASE WHEN LOWER(CAST(r.comments AS VARCHAR (MAX))) LIKE '%great%' 
OR LOWER(CAST(r.comments AS VARCHAR (MAX))) LIKE '%comfortable%' 
OR LOWER(CAST(r.comments AS VARCHAR (MAX))) LIKE '%good%' THEN 1 ELSE 0 END) / COUNT(r.ID)) > 0.5;

5. Revenue by Host for Recent Reviews

Calculated revenue from listings reviewed in the last year to understand recent host performance.

Result: Lists hosts with notable recent revenue performance.


SELECT 
    l.host_name,
    SUM(p.price) AS total_revenue
FROM listings l
JOIN pricing p ON l.id = p.listing_id
JOIN reviews r ON l.id = r.listing_id
WHERE r.date IS NOT NULL
AND r.date >= DATEADD(YEAR, -1, GETDATE())
GROUP BY l.host_name;

6. Hosts Named "Stan"

I performed a case-insensitive search to find hosts named Stan.

Result: Provides a count of hosts named Stan, useful for networking or demographic analysis.


SELECT 
    DISTINCT l.host_name
FROM listings l
WHERE LOWER(l.host_name) = 'stan';

7. Total Revenue from High-Priced Listings

Revenue generated by listings priced above $700 helps identify high-value properties.

Result: Shows the revenue impact of high-priced listings.


SELECT 
    l.id AS listing_id,
    l.name AS listing_name,
    SUM(p.price) AS total_revenue
FROM listings l
JOIN pricing p ON l.id = p.listing_id
WHERE p.price > 700
GROUP BY l.id, l.name;

8. Most Reviewed Listing

This query finds the listing with the highest review count.

Result: Highlights the most popular listing based on review volume.


SELECT TOP 1
    l.id AS listing_id,
    l.name AS listing_name,
    COUNT(r.ID) AS review_count
FROM listings l
JOIN reviews r ON l.id = r.listing_id
GROUP BY l.id, l.name

9. Listings Reviewed in Q4 2022

Finds listings with reviews from October to December 2022, focusing on recent performance.

Result: Provides data on recent listings’ review activity.


SELECT 
    l.id AS listing_id,
    l.name AS listing_name,
    r.date AS last_review_date
FROM listings l
JOIN reviews r ON l.id = r.listing_id
WHERE r.date IS NOT NULL
AND r.date BETWEEN '2022-10-01' AND '2022-12-31';

10. Categorization of Listings as High- or Low-Priced

Using a neighborhood average, I categorized listings into "high-priced" or "low-priced."

Result: Shows the categorization of listings and highlights the potential pricing strategy.


    WITH neighborhood_avg AS (
    SELECT 
        neighbourhood,
        AVG(p.price) AS avg_neighborhood_price
    FROM listings l
    JOIN pricing p ON l.id = p.listing_id
    GROUP BY neighbourhood
)
SELECT 
    l.id AS listing_id,
    l.name AS listing_name,
    p.price,
    CASE 
        WHEN p.price > na.avg_neighborhood_price THEN 'high-priced'
        ELSE 'low-priced'
    END AS price_category,
    SUM(p.price) AS total_revenue
FROM listings l
JOIN pricing p ON l.id = p.listing_id
JOIN neighborhood_avg na ON l.neighbourhood = na.neighbourhood
GROUP BY l.id, p.price, na.avg_neighborhood_price, l.name;

Conclusion

This analysis offers insights into Seattle's Airbnb market by examining revenue, pricing, guest sentiment, and listing performance. These findings provide Airbnb hosts with actionable insights for optimizing pricing and improving guest experience.

Recommendations

  1. Dynamic Pricing: Adjust prices seasonally to match occupancy patterns.
  2. Customer Satisfaction: Encourage positive feedback by enhancing amenities.
  3. Targeted Marketing: Promote high-rated and high-priced listings in premium areas.